Workforce planning is one of the most complex tasks to capture in software. It needs to compose shifts, track time registrations, and manage flexible staff. Each component is a fully-fledged application in its own right.
Intus Workforce Solutions brought all of this together in InPlanning: a complete system covering the full spectrum of workforce planning. This doesn’t just deliver operational convenience, but also generates large amounts of data. Data with far more potential than what had been tapped until now.
The Challenge
Using Analytics, Intus was already making management information widely available; the vast majority of customers used it to access data warehouse insights. But this approach was running into its limits.
The management information Intus offered was incomplete, difficult to scale, and hard to extend. On top of that, customers could only access this data through their own data warehouse. InPlanning users themselves had no direct access to data aligned with their own authorization level.
Intus also wanted to scale InPlanning’s capacity. Using Kubernetes in the cloud allows deployments to be spun up and scaled quickly. The only problem? Data sovereignty. Even in Kubernetes, customer data sits on third-party servers.
Intus’ customers in the healthcare and (semi-)public sector have a strict requirement that their data stays in Europe and does not end up with American hyperscalers such as AWS, Azure, or Google Cloud. For Intus, this isn’t a technical concession - it’s a deliberate strategic choice.
The vision is that cybersecurity, privacy, and data sovereignty come first in InPlanning. This translates into investment in talent that builds infrastructure on open-source software capable of running on any European server or cloud.
The Solution
Wolk developed an analytics data platform for InPlanning together with Intus, built entirely on open-source software. The platform transforms InPlanning data into usable, actionable insights - whether that’s in a Business Intelligence dashboard, or made available as datasets for data scientists at Intus’ customers.
The Insights Data Platform is built from the following open-source components:
- ClickHouse - a columnar, OLAP-based database engine for replicating the InPlanning application database and running data transformations.
- SQLMesh - a Python-based tool for data transformations. The replicated InPlanning data is periodically transformed using a Medallion Architecture, producing a star schema for analysis - think fact tables for shifts and dimension tables for employees.
- Apache Superset - a Business Intelligence dashboard application. In Superset, data is aggregated for visualizations and charts, developed by Intus and its customers.
- OVHcloud - a French hyperscaler with a mature Managed Kubernetes Service where the data platform’s Helm charts run.
A critical aspect of this architecture: each customer runs on their own dedicated data stack - their own ClickHouse instance, their own transformation pipeline, their own dashboard environment, fully isolated from other customers. This also makes the platform flexible: you can start with a small Kubernetes footprint and scale up as the terabytes start flowing in.
The Collaboration
Building an open-source data platform requires balancing two types of knowledge: specialization in DevOps and data engineering, and deep domain expertise in InPlanning and how customers use it to manage schedules for thousands of employees. This balance defined how the development sprints with Intus’ developers were run.
To prototype quickly, Wolk and Intus developed the SQLMesh data models and the Helm charts simultaneously, shaping the data stack in parallel. This allowed rapid experimentation with the data transformation flow and the BI insights that could be built from InPlanning data. It also meant we could present early demo environments to Intus’ customers, creating a direct feedback loop with future users.
“The project started with one big question: how do we make InPlanning data truly useful for our customers? What I value most about working with Wolk is that we didn’t just build a technical platform - together we constantly weighed what’s feasible against what customers actually need. The Insights product we have today is proof of that.” - Sarwin, Product Owner Insights @ Intus
The Result
A select group of Insights Data Platform beta customers can now build their own self-service BI dashboards using near-real-time planning data. They can also work with the raw data: the full schema can be imported from a dedicated S3-based bucket, allowing customers to explore the insights most relevant to their own planning situations.
For Intus, this means less pressure on technical consultants. Standard dashboards or analysis tables can now be deployed centrally in one place. A consultant can gather requirements from one customer and roll them out to all customers - serving not just one use case, but bringing new insights to every InPlanning user.
The Future
A number of Intus customers have already been onboarded onto the data platform beta. The remaining customers will follow later this year, at which point internationalization will also be required for Intus’ customers abroad.
The data platform itself is the foundation on which further development will continue. With all planning data in one place, the possibilities extend beyond dashboards to predictive algorithms and deeper analyses.
Data science hypotheses like “given current turnover rates, where will we run into staffing shortages this year?” become viable. The ultimate goal is to use these deeper analyses to uncover more fundamental challenges at Intus’ customers, making their planning processes even more effective.
Want to turn historical data into actionable insights? At Wolk, we help organizations design and build data platforms tailored to their specific requirements. Get in touch at info@wolk.work.
